Skip to main content

Belgium's ISI combats Carousel Fraud with Advanced Analytics

· 3 min read
Michael Thompson
Regulatory Enforcement Officer

By Jérôme Bryssinck

Introduction​

The Special Tax Inspectorate (ISI) in Belgium, a division under the Federal Public Service Finance, has successfully combated carousel fraud in the country utilizing advanced analytics from SAS. Yannic Hulot, the ISI's director, reveals the success story.

Carousel fraud, a form of Value-Added Tax (VAT) theft, allows criminal organizations to take advantage of the weaknesses in the VAT system and misappropriate vast amounts of VAT. The fraud involves a criminal selling goods to a conspirator in another country through a business-to-business sale without paying VAT to the government. The accomplice then sells the goods, collects VAT, and shows no intention of paying it back to the authorities. This cycle continues, leading to significant losses for the government.

In Belgium, losses from VAT fraud amounted to €1.1 billion in 2002. Following the introduction of the hybrid detection model, losses decreased to a negligible €18.5 million in 2012 - a reduction of 98 percent.

Our core business is to investigate organized and serious fraud, especially those with complex mechanisms and an international dimension. We require fairly advanced risk analysis, as our methods are not for mass detection. We look for the very essence of the fraud – persons with certain characteristics, engaged in suspicious transactions, and with information from abroad. -Yannic Hulot, Director, ISI

The ISI focuses on the detection and prevention of sophisticated fraud schemes with international connections. Managing and verifying the extensive data is a significant challenge, especially with the increasing amounts of data from emerging events like the European Savings Directive and foreign collaborations. SAS hybrid detection model has proven effective in handling the data processing and analysis.

The Importance of a Similar Solution in Europe​

The European Union (EU) faces a substantial estimated VAT gap of nearly €100 billion per year, which includes carousel fraud, irregularities, errors, and classic fraud. Recognizing the success of Belgium's implementation, other countries are looking for similar solutions to tackle VAT fraud.

Effectiveness of the Hybrid Detection Model​

The SAS hybrid detection model is an innovative technique that utilizes multiple analytical methods to expose hard-to-find fraudsters. With the success in Belgium, this detection method is increasingly regarded as the go-to solution for VAT fraud detection throughout Europe.

A VAT carousel represents a network, making social network analysis an ideal tool for detection. The hybrid detection model, which leverages social network analysis, is indispensable for extracting relevant information from the vast amounts of data and identifying the connections essential for uncovering carousel fraud.

Expanding the Success in Belgium​

The number of ISI specialists has increased from 500 to 600. Presently, all fraud investigators have access to this powerful data mining tool. The future outlook for Europe is positive, with Yannic Hulot expressing his hope for a common cross-border analysis approach that can save substantial amounts for European taxpayers.